Detecting pattern-based outliers

被引：32

作者：

Hu, TM ^{[1
]}

Sung, SY ^{[1
]}

机构：

[1] Natl Univ Singapore, Dept Comp Sci, Singapore 117543, Singapore

来源：

PATTERN RECOGNITION LETTERS | 2003年 / 24卷 / 16期

关键词：

outlier detection; complete spatial randomness; clustering; regular spacing;

D O I：

10.1016/S0167-8655(03)00165-X

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Outlier detection targets those exceptional data that deviate from the general pattern. Besides high density clustering, there is another pattern called low density regularity. Thus, there are two types of outliers w.r.t. them. We propose two techniques: one to identify the two patterns and the other to detect the corresponding outliers. (C) 2003 Elsevier B.V. All rights reserved.

引用

页码：3059 / 3068

页数：10

共 14 条

[1]

Aggarwal C. C., 2001, SIGMOD Record, V30, P37, DOI 10.1145/376284.375668

[2]

BARNETT V, 1994, OUTLINERS STAT DATA

[3] LOF: Identifying density-based local outliers [J].

Breunig, MM ;

Kriegel, HP ;

Ng, RT ;

Sander, J .

SIGMOD RECORD, 2000, 29 (02) :93-104

[4]

Cressie N, 1993, STAT SPATIAL DATA

[5]

DIGGLE P, 1985, J R STAT SOC C-APPL, V34, P138

[6]

Ester M., 1996, Proc. Second Int. Conf. Knowl. Discov. Data Min, P226, DOI DOI 10.5555/3001460.3001507

[7]

GRANVILLE V, 1995, J ROY STAT SOC B MET, V57, P501

[8]

Hawkins D., 1980, Identification of Outliers, DOI DOI 10.1007/978-94-015-3994-4

[9] Distance-based outliers: algorithms and applications [J].

Knorr, EM ;

Ng, RT ;

Tucakov, V .

VLDB JOURNAL, 2000, 8 (3-4) :237-253

[10]

Murphy P.M., 1994, UCI REPOSITORY MACHI

← 1 2 →